Intramolecular Photocyclization of 2-Vinylbiphenyl-Like Compounds. 2

(16) Bergmann, E.; Bondi, A. Ber., 1931, 64, 1455-1480. (17) Dickerman. S. C.: Feigenbaum, W. M. J. Chem. Soc., Chem. Commun.,. (18) Hatchard, C. G.; ...
1 downloads 0 Views 540KB Size
Fornier de Violet et al.

/ Photocyclization of 2- Vinylbiphenyl-Like Compounds (22)

R. J. Am. Chem. Soc., following paper in this issue. Patterson, L. K.; Porter, G.; Topp, M. R. Chem. Phys. Lett., 1970, 7,

(23)

Rabideau, P. W.; Harvey, R. G.; Stothers, J. B. J . Chem. Soc., Chem.

(24)

Hixson, S . S.; Mariano, P. S.; Zimmerman, H. S. Chem. Rev., 1973, 73,

Bergmann, E.; Bondi, A . Ber., 1931, 64, 1455-1480. Dickerman. S. C.: Feigenbaum, W . M. J. Chem. Soc., Chem. Commun.,

(25)

Parker, C. A. Photoluminescence of Solutions, Elsevier: Amsterdam, 1968;

1966,345-346.

(26) (27)

Wagner, P. J. J. Am. Chem. Soc., 1967, 89, 2820-2825. Swenton, J. S.; Ikeler, Th. J.; Le Roy Smyser. G. J . Org. Chem., 1973,38.

(28) (29) (30)

Swenton. I. S.; Leroy Smyser, G. J. Org. Chem., 1978, 43, 165-167. Imamura, A.; Hoffmann. R. J. Am. Chem. SOC.,1968, 90, 5379-5385. Bertini, F.; Grasseli, P.; Zubiani, G.; Cainelli, G. Tetrahedron, 1970, 26,

(!O) Bradsher, C. K. J. Am. Chem. Soc., 1944, 66, 45-50. ( '! Bachmann, W. E.; Brockway, C. E. J. Org. Chem., 1948, 13,384-389. (12) Ciois, P. H. J. Org. Chem., 1957, 22, 135-138. (13) Haworth, R. D. J. Chem. Soc., 1932, 1125-1133. (14) Martin, E. L. J. Am. Chem. SOC..-1936, 58, 1438-1442. (15) Rutherford, K. G.; Newman. M. S. J. Am. Chem. Soc., 1957, 79, 213N

61 2-614.

Commvfl., 1969, 1005-1006. 531-551.

214.

(16) (17) (18)

Hatchard, C. G.; Parker, C. A. Proc. R. Soc, London, Ser. A, 1956, 235,

p 314.

518-536. (19) (20)

Bowman, W. D.; Demas, J. N. J. Phys. Chem., 1976,80, 2434-2435. Singer, L. A.; Brown, R. E.; Davis, G. A. J. Am. Chem. Soc., 1973, 95, 8638-8643.

(21)

6683

Fornier de Violet, P.; Bonneau, R.; Lapouyade, R.; Koussini, R.; Ware, W.

1157-1 166.

128 1- 1290.

Intramolecular Photocyclization of 2-Vinylbiphenyl-Like Compounds. 2. Detection of the Intermediates and Kinetic Study by Laser Flash Photolysis of 1-(o-Diphenyl)- 1-phenylethylene Ph. Fornier de Violet,*' R. Bonneau,' R. Lapouyade,' R. Koussini,' and W. R. Ware Contribution from the Laboratoire de Chimie Physique A, Uniuersit; de Bordeaux I , 33405 Talence, France. Received March 21, I978

Abstract: Static fluorescence measurements and laser flash photolysis experiments have permitted the elucidation of the mechanism of photocyclization of 1 -(o-diphenyl)- 1 -phenylethylene (a 1 , l -diarylethylene). By fluorescence measurements, the quantum yield and the lifetime of fluorescence are found to be $q= 0.09 and T f = 4 X s-I, respectively. By flash photolysis, two transients species having different lifetimes have been observed. One, whose absorption has a maximum a t 370 nm (e,, (1.7 f 0.3) X IO4 M-l cm-I), with a decay time on the order of 1 p s , is scavenged by oxygen ( k =~1 .O X IO9 M-I s-I). It has been assigned to the triplet state. The second one, absorbing in the visible ,A,(, 510 nm, tmax (1.2 f 0.2) X IO4 M-l a n - [ ) and below 300 nm, has a lifetime of the order of 1 ms. I t is scavenged by oxidizing agents such as iodine ( k i , = 4.5 X I O h M-I s-I) and is attributed to a cyclic intermediate. By both continuous irradiation and flash photolysis, we find that the cyclic intermediate is formed from the singlet and triplet excited states, in the ratio 4.1. Rate constants as well as conversion yields of the main elementary reactions have been determined.

I. Introduction

Scheme I1

The oxidizing photocyclization of hexatrienes having one or more double bonds engaged in an aromatic system has been extensively studied.2 In the case where two consecutive double bonds are part of phenyl rings such as in 1, the UV irradiation in a nonoxidizing media leads to a stable cyclic product which would be formed by a light-induced cyclization of the disguised hexatriene system, followed by a thermally allowed suprafacial 1,5 hydrogen ~ h i f t . ~ - ~ The intermediate species postulated in this mechanism has never been observed. W e report here evidence for this intermediate obtained by a laser flash photolysis study of the photocyclization of a 1 ,I-diarylethylene. A kinetic scheme is presented which accounts for the rates and products of the photoreaction. W e have chosen to study 1-(o-diphenyl)-1-phenylethylene (DPE), which gives under UV irradiation 9-phenyl-9,lOdihydrophenanthrene ( P I ) in nonoxidizing media, while 9phenylphenanthrene (P2) is formed in the presence of oxidizing agents. Scheme I

0002-7863/78/1500-6683$01.00/0

*Ph

11. Experimental Section 1. Materials. The preparation of 1 -(o-diphenyl)- 1 -phenylethylene has been described elsewhere.2 2. Fluorescence Measurements. Fluorescence decay curves were measured using the time-correlated single photon counting technique (PRA system).6 Fluorescence spectra and quantum yield of corrected fluorescence were measured with a spectrofluorimeter (Perkin-Elmer MPF.3). The solutions were unstable against prolonged exposure to the arc of the spectrofluorimeter; this effect was minimized by using very narrow slits and high gain. 3. Transient Absorption Measurements. The transient absorption measurements were made by laser photolysis using the fourth harmonic a t 265 nm of a Q-switched neodynium laser (Quantel instrument) delivering approximately 80 mJ of radiation in 3 ns. The analyzing beam from a pulsed 75-W xenon lamp was perpendicular to the laser beam and passed through a quartz cell (10 mm square, all four sides polished) containing the sample. The apparatus and the techniques have been described elsewhere.' 0 1978 American Chemical Society

6684

Journal o f t h e American Chemical Society

/

100:21

/

October 11, 1978

Table I. Comparison of Some Spectral Properties of Diphenyl- and I -(o-Diphenyl)- 1-phenylethylene in Cyclohexane as Solvent absorption compd

0-0

T

emission e,

hn‘l,,

nm

I O4 M-’ cm-’

nm

@F

decay time T , ns

247

1.5

0.18

16.0

245

2.8

305 318 310 322

0.09

4.0

~lnax.

Ph

Table 11. Oxygen Effect on the Lifetime of the Transient Species Absorbing at 510 and 370 nm absorption at 510 nm 1/T = Z k f k ~ [ O 2 ]’ ,

f

LO21

T,

1

solution

M

ms

S-‘

degassed aerated 02-saturated

-2.3 x 10-3 -1.1 x 10-2

6.1 4.40 2.40

I64 225 417

AAbsorbonce -0,

d q

0.10.

p‘

A / ‘0 ‘0

Q05-

)I

\!

At \

\

q “/

,

\

io

,

0

0

\

\

*\&

111. Results

Quantum Yield and Lifetime of Fluorescence. The absorption and fluorescence spectra of DPE in cyclohexane solutions are quite similar to those recorded for biphenyl in the same solvent8 (see Table I) except for a red shift of the fluorescence spectra. The quantum yield of fluorescence &of DPE was determined by comparison of its intensity of fluorescence with the case of biphenyl for which @ f is known (@(= 0.18) under the same experimental condition. It was found to be 4,- = 0.09. The lifetime of fluorescence of DPE in degassed solution (3 X M in cyclohexane) obtained by the single photon counting technique was determined using 306-nm excitation while monitoring emission a t 325 nm. Using an iterative convolution by a least-squares method,6 a single exponential fluorescence decay was obtained, giving a lifetime of 4.0 ns at 25

OC. Transient Absorption Measurements. By laser flash photolysis of DPE in degassed cyclohexane solutions, two transient absorptions, distinguishable by their decay time, were observed between 290 and 600 nm. (1) The first intermediate, having a “long decay time” in the millisecond range, appeared in the red (A,, 510 nm) and in the UV (A,, < 300 nm). The species that gives rise to this absorption disappears by a first-order process with a rate constant of 1.60 X lo2 s-l. Measurements at wavelengths shorter than 300 nm are difficult and rather inaccurate since

absorption at 370 nm I/. = zk kQ[z], ’

I

+

T

1.75 FS 400 ns 87 ns

x 10-5

s-’

5.7 25 115

the fluorescence emission and the permanent absorption due to the photoproduct interfere with one another. (2) The second intermediate, having a decay time in the microsecond range, appeared between 340 and 440 nm with a maximum a t 370 nm. This transient absorption also disappears by a first-order process with a rate constant of 5.75 X I O 5 s-l in degassed solutions (see Figure 1). To identify the nature of these transient species and their role in the kinetics of the reaction scheme, further experiments were performed. Oxygen Effect. By photolysis in continuous illumination,* it was shown that the UV irradiation of DPE leads to a cyclic nonoxidative product, 9-phenyl-9,lO-dihydrophenanthrene (PI), either in degassed or aerated cyclohexane solution. The total quantum yield was found to be slightly but significantly

spectral region. W e looked for an effect of oxygen in the decay kinetics of the transient absorptions in degassed, aerated, and 02-saturated solutions. It was found that oxygen has a small influence on the decay of the species having “a long lifetime” but is a very efficient scavenger of the short lifetime transient species (see Table 11). From the plot of the reciprocal measured lifetimes as a function of 0 2 concentration (1 / 7 = 1/ 7 0 -k k ~ [ o z the ]) value of the rate constant for oxygen quenching, kQ, was found to be equal to 2.3 X IO4 and 1 .O X IO9 M-l s-l, respectively, for the long- and short-lived transient species. From examination of the effects of oxygen on the formation of the transient absorption, the following observations can be made. (1) The transient absorption at 370 nm is formed during the laser pulse duration (3 ns) whatever the 0 2 concentration. (2) In either 02-saturated or aerated solution, the species absorbing a t 510 nm is also produced during the laser pulse; in the degassed solution the optical density still increases slowly during the dark period following the irradiation. This can be seen in Figure 2 where O D at 510 nm is plotted as a function of the time. It can be seen that the contribution of the slowly growing transient absorption is approximately 20% of the entire transient absorption. Plotting In (OD,,, - OD,) as a function of time, we found that this slow process of formation follows first-order kinetics with a rate constant k = 5 X lo5 s-l which is quite close to the

Fornier de Violet et al.

/

Photocyclization of 2- Vinylbiphenyl-Like Compounds

0

Figure 2. Transient optical density a t 510 nm vs. time: (1) in degassed solutions; ( 2 ) in aerated or 02-saturated solutions; (3) In (OD,,, - OD,) vs. time.

T

4(r')

";ol

/ /

1

6685

5

Figure 4. Reciprocal maximum concentration of 9-phenylphenanthrene against the reciprocal concentration of 12.

/

I /

v

1CCU

I

1

f+=C

-Ph

PI

D.P.E

2

3

4

5

6 IIdxlOflM

p2

Figure 5. Schematic diagram of the oxidizing and nonoxidizing photocyclization of 1-(o-diphenyl)- I -phenylethylene.

Figure 3. Reciprocal lifetime of the transient absorption at 510 nm as a function of the concentration of 12.

value found for the decay of the transient absorption band a t 370 nm. Iodine Effect. In the preceding paper it was shown that in the presence of iodine, the continuous irradiation of DPE leads to an oxidizing cyclization giving the product P2 as a new final product a t the expense of the nonoxidizing process giving

PI. In laser flash photolysis experiments, when the iodine concentration is gradually increased up to 5.5 X M, the decay time of the short-lived transient (370 nm) is not significantly changed while that of the long-lived transient (31 5, 510 nm) decreases strongly. From the slope of the straight lines obtained by plotting the reciprocal decay time a t 510 and 31 5 nm as a function of I2 concentration (Figure 3) the rate constant of the reaction of iodine with the long-lived transient is found to be k l , = 4.5 X lo6 M-' s-'. As the iodine concentration increases a residual permanent absorption appears at 3 15 nrn after the decay of the transient absorption related to the long-lived species. This permanent absorption is due to the 9-phenylphenanthrene, P2, the extinction coefficient of which is around 1800 a t 3 15 nm.9 This allows evaluation of the amount of P2 formed in a single laser pulse for a given 12 concentration. This amount of P2 increases as the 12 concentration increases but tends to a limitive value for [ I 2 1 > 5 X M. The ordinate intercept of the straight line obtained by plotting 1 /[P2] vs. l/[Iz] gives the maximum

concentration of P2 which could be found, [P21max= (2.2 f M (see Figure 4). 0.4) X 111. Interpretation

The experimental data described above can be given a consistent interpretation, assuming the mechanism of Scheme 111 (see Figure 5). Scheme Ill hu

So-SI

6686

Journal of the American Chemical Society

In this mechanism, one must note that the cyclic intermediate product X could be formed in steps 5 and 6, via the singlet and triplet pathways. It has also been considered that the species X can give back the ground state by reversible opening (step 1 1 ) as has been shown to be the case for the similar molecule, 1,2-diar~lethylene.~-~ Assignment of the Transient Absorptions. The decrease of the lifetime of the transient species absorbing a t A,, 370 nm, as the 0 2 concentration is increased, suggests that this species should be assigned to the triplet state. The unusually short lifetime (-1.7 ps) in degassed medium is not surprising if one considers that the triplet state contributes to the photocyclization, via the intermediate X. The species having a long lifetime, with a maximum of absorption a t 5 I O nm and below 300 nm, is attributed to the intermediate X. This is produced by an intramolecular cyclization from both the singlet and triplet state, since it is formed both during the laser pulse duration and in the dark period following excitation. The mechanism of formation of X is confirmed if one considers the following. ( I ) The lifetime of the singlet state is 16 ns in the case of biphenyl while it is only of the order of 4 ns in the case of DPE, implying the existence of an extra deactivation pathway; this additional deactivation process can be attributed to the cyclization process. (2) The decay of the triplet state and the slow formation of X have approximately the same rate constant. Evaluation of the Extinction Coefficient of the Transient Bands. In order to evaluate the extinction coefficient of a transient species in a flash photolysis experiment one must know the absorbance of the transient and the concentration of the species. The concentration of X formed by a single laser pulse can be deduced from that of the final products P I and P2 whose ratio depends on the experimental conditions, if it is assumed that the conversion of X in PI or P2 is quantitative; the concentration of the stable products P I and P2 can be measured by straightforward colorimetry. However, if a reversible opening of X (reaction 1) occurs, the concentration of X could be underestimated. Combined experiments by continuous irradiation and laser flash photolysis show that the reaction 1 1 is inefficient compared to reactions 10 12. Indeed, in continuous irradiation the quantum yields of product formation in degassed or in aerated solutions containing iodine at 5 X M are almost identical: 4 p , = 0.24 and 4 p , + p 2= 0.04 0.22 = 0.26. From laser photolysis measurements it was found that addition of this amount of iodine decreases the lifetime of X by a factor of 20. The yield of X So should be similarly reduced. If the slight increase in quantum yield noted above were significant, it could be due to a reversible opening. The effect, however, must be small, since on the basis of these data, the yield of conversion of X to products is at least 0.92. By measuring [PI] or [P2] by the change of the optical density a t 300 or 3 15 nm, respectively, the extinction coefficient of X has been deduced to be ~ 5 1 0 =, ~(1.1 f 0.2) X IO4 M-I cm-l. The extinction coefficient of the triplet species a t 370 nm is evaluated as follows. From laser photolysis measurements, we found that about 20% of the X species is produced through the triplet pathway. Thus the concentration of triplet produced by a single laser pulse is given by 0.2[X] = @TP~[T] and with [XI = (2.2 f 0.4) X M , [TI = (7.6 f 1.5) X M leading to a n extinction coefficient value q 7 0 , m of (1.7 f 0.3) X IO4 M-I cm-l. The limited accuracy of the ~ s and~€370nm o values ~ is~related to the difficulty of knowing PI and Pz concentrations. The accuracy of the e300 ( P I ) and €315 (Pz) are similarly limited, since the wavelengths 300 and 3 15 nm are on the steeply rising side of the absorption bands of PI and P2.

+

+

-

/

100:21

/

October 11, I978

Table 111. Values of the Lifetimes, Rate Constants, and Yields Measured or Deduced in the Photocvclization of DPE

Evaluation of the Rate Constants and of the Internal Conversion Yield. Data obtained by fluorescence, flash photolysis, and continuous irradiation measurements are summarized below. 7f

= I/(kf

+ k,,, + kST + ksx) = 4 X @f

TT

(1)

s

= sfkr = 9 X

+ k , + k,,,) = 1.75 X TX = l / ( k x p + kxs,,) = 6.1 X

(11)

= l/(kTx

s s

~ x =40.58 ~ ~ ~ 4 ' s ~=~4 ~ x +4~ S~ T ~~ T ~X ~=X0.24 P ~

(111) (IV)

~ T =P 4

(VI

(VI)

= 4 S X 4 X P l = 0.21

(VII) when the triplet state is quenched by piperylene From this set of results, the following rate constant can be deduced: 4SPI

k f = $ f / T f = 2.25 X l o 7 s-' Since the phosphorescence lifetime is a few seconds a t 77 K2 the rate constant of phosphorescence emission k, is smaller than 1 s-I and can be neglected in comparison with other processes occurring a t room temperature. From the triplet lifetimes measured in flash photolysis we found that k T X k,,, = 5.7 X IO5 s-l. Since 4xpI is equal to or very near unity (4xpI > 0.92) we have directly k x p , = 1.64 X IO2 s-l and from eq v, 4 ~ =x 0.58 = k T X r t and thus k T X = 3.3 X lo5 s-I; from eq V I I , 4 s =~ k s X T f N 0.21 and thus ksx = 5.2 X lo7 s-l; from eq VI and VI1 &,T+TX = 0.03 and with 4 ~ =x 0.58, we obtain 4 s =~ 0.05 = kSTrf ( k S T = 1.3 X I O 7 s-l); from eq 111 giving kTx k,,, and from eq V giving kTx, we have k,,, = 2.4 X IO5 s-l; from eq I and the calculated values of kf, ksr, ksx, we have k,,, = 1.6 X los s-l. It is important to note that, although the number of parameters involved in the proposed mechanism is large, many different methods of measurements have been used to separate the rate processes involved. The accuracy of the rate constants and of the quantum yields of conversion deduced from primary data have errors which we estimate to be f10 %. The accuracy of rate constants calculated from differences between experimental measurements is more limited, as, for instance, is the case for the value of kST which arises from a small difference between 4' and 4 (eq V I and VII). Taking into account an accuracy of 5% on 4' and 4, we estimate 0.01 < 4~~6-j-x < 0.06 and thus 4.3 X IO6 s-I < ksT < 2.6 X lo7 s-l. A more precise value of ksT can be obtained from flash photolysis measurements, since 20% of X is formed via the triplet pathway. That implies that 4 s N ~ &$sT+TX and with the values of 4 s (0.21 ~ and &-x (0.58) this gives ST 0.1 and thus kST = 2.5 X l o 7 s-l. All the measured and calculated results are collected in Table 111.

+

+

-

IV. Conclusion A kinetic scheme for the photocyclization of 1-(0-diphe-

Michl et al.

1 Photochemical Conuersion of a Polycyclic Cyclobutene

ny1)- 1-phenylethylene is proposed based on product analysis, fluorescence measurements, and transient species detected by flash photolysis experiments. The rate of photocyclization from the singlet state is 100 times faster than the apparent cyclization rate constant from the triplet state. This can be accounted for in the following manner. (1) The intermediate X, being a ground-state singlet species, can be formed more easily from an excited state of the same multiplicity and probably via an electrocyclic ring-closure route. (2) The energetic factors can be important. I f one assumes the existence of a common transition state from the excited singlet and triplet states to produce X, it would follow that the activation energy would be smaller for reaction from the singlet state than from the triplet state. Hence photocyclization from the singlet state would be favored kinetically. (3) It is more likely, according to the results described in part 1,2 that the mechanisms of formation of X via the singlet and triplet state would be different, since the twisting of the double bond will differ in these two states affecting the spatial arrangement of the atoms to be linked.1° Further experiments will be carried out with other diarylethylenes.

Acknowledgments. The authors wish to thank Professor J . Joussot-Dubien for his interest in this work and helpful com-

6687

ments. One of us (P.F.V.) is grateful to the National Research Council of Canada and the Centre National de la Recherche Scientifique for financial support of this work.

References and Notes (1)Equipe de Recherche Associee au CNRS no. 167. (2)See part 1: Lapouyade, R.; Koussini, R.; Fornier de Violet, P. J. Am. Chem. SOC.,preceding paper in this issue.

(3)Horgan, S. W.; Morgan D. D.; Orchin, M. J. Org. Chem., 1973, 38, 3801-3803. (4) Laarhoven. W. H.; Cuppen, Th. J. H. J. Chem. Soc., Perkin Trans. 1, 1972, 2074-2079. (5)Lapouyade, R.; Koussini, R.; Rayez, J. C. J. Chem. Soc., Chem. Commun., 1975,675-676. ( 6 ) Ware, W. R.; Doemeney. L. J.; Nemzek, T. L. J. Phys. Chem., 1973,77. 2038-2048. (7)Fornier de Violet, P.; Bonneau R.; Joussot-Dubien, J. Chem. Phys. Lett., 1974,28,569-572. (8)Berlman. I. B. Handbook of Fluorescence Spectra of Aromatic Molecules, Academic Press: New York, N.Y., 1955;p 92. (9) At this wavelength, which is on the steeply rising side of the absorption band of P2, the slits of the monochromator are around 0.5mm, giving a resolution of 3 mm. It follows that f315nm = 1800 z t 350 M-' cm-'. (10) The values of the rate constants kTxand knrghave been evaluated by assuming that the triplet state observed by flash spectroscopy is the cyclizing species. On account of its lifetime, this triplet should be a "planar" state, the ethylenic CH2 still being in the molecular plane. On the contrary, if the cyclization of the triplet occurs after rotation of the double bond, on one hand, the lifetime of the observed triplet would give the rate of rotation of the double bond and not that of cyclization: on the other hand, the rate of cyclization of the "twisted" triplet should be much larger than the 3.3X lo5 s value reported above, since its lifetime must be much shorter than

1.75ps.

r,r-Biradicaloid Hydrocarbons. The Pleiadene Family. 4.' Involvement of Upper and/or Hot Excited Singlet and Triplet States in a Photochemical Conversion of a Polycyclic Cyclobutene to a Butadiene Alain Castellan,* Jaroslav Kolc, and Josef Michl* Contribution from the Department of Chemistry, Unicersity of Utah, Salt Lake City, Utah 841 12. ReceiGed April 24, 1978

Abstract: The mechanism of the photochemical electrocyclic transformation of the cyclobutene 2 to pleiadene 1 in a rigid glass (77 K ) with long-wavelength light involves successive absorption of two photons, with triplet 2 as the intermediate state. At shorter wavelengths, the reaction proceeds by the ordinary one-photon mechanism. It is concluded that thermalized SI and T I states a r e unreactive and that the reaction occurs in upper and/or hot excited states with quantum yields of the order of 1%. It is pointed out that the common practice of ignoring such possible upper and/or hot excited state contributions to unimolecular photochemical reactivity of organic molecules in solution is questionable.

The photochemical preparation of pleiadene (1) by irradiation of the cyclobutene precursor 2 in rigid organic glasses has been rep0rted.39~Alkyl derivativeP as well as more highly condensed polycyclic analogues4s7 are similarly accessible. We now wish to report in detail our mechanistic investigation of this photochemical cyclobutene butadiene conversion. Its unusual nature has already been briefly d e ~ c r i b e d ~ .and ' - ~ the theoretical implications were discussed.

-

Results Excited States of 1 and 2. The absorption spectrum of 1 and seven of its alkyl derivatives was reported and analyzed prev i o u ~ l yThe . ~ ~energies, ~ polarizations, and relative intensities of the first six transitions were in excellent agreement with velectron calculations, which led to an assignment of the second excited singlet state as doubly excited. The first tran0002-7863/78/ 1500-6687$01 .OO/O

sition is short-axis polarized and lies in the near-IR region (0-0 band a t 1 1 700 cm-I); no fluorescence was observed. The lowest triplet is calculated to lie only about 3000 cm-' above the singlet ground state and no experimental evidence for it has been obtained. In a rigid glass a t 77 K, 1 is stable for many hours, both in the dark and under UV irradiation. The absorption and uncorrected fluorescence and phosphorescence spectra of 2 are shown in Figure 1 and correspond well to expectations for noninteracting acenaphthene and benzocyclobutene chromophores. The absorption reveals the expected presence of the three low-energy transitions of the naphthalene chromophore (N1 = ILL,,N 2 = ILa, N 3 = I B b ) and of the lowest benzene transition (B1 = ' L b ) . Polarization measurements on these transitionsI3 by the stretched polymer method are in perfect agreement with expectations, given the C, symmetry and the 1 15' dihedral angle between the plane

0 1978 American Chemical Society